During embryogenesis, the vertebrate central nervous system (CNS) becomes subdivided (patterned) into several structures. One of these, the hindbrain, gives rise to the cerebellum and brainstem, which regulate basic functions (e.g., posture and movement) and cognitive functions (e.g., language). Hindbrain defects are therefore associated with both motor control problems (e.g., ataxias) and cognitive defects (e.g., autism). To better understand neural patterning in general and hindbrain development in particular, we recently cloned several genes expressed early during hindbrain development, including the zebrafish meis3 gene. Based on our results to date, we hypothesize that Meis3 acts as a Hox cofactor during hindbrain development by forming transcription regulatory complexes with Hox and Pbx proteins. Specifically we find that Meis proteins form trimeric complexes with Pbx and Hox in vitro and that Hox proteins are unable to activate transcription in the absence of Pbx and Meis in vivo. However, several questions remain unanswered. First, it is unclear if Meis, Pbx and Hox proteins always function together, or if they also have independent functions. To address this question we will first compare the effects of reducing Meis, Pbx and Hox function in vivo. This will both demonstrate whether the proteins act together (i.e., reducing their levels give the same phenotype) and reveal the role for each protein during hindbrain development. We will then directly address whether the three proteins must interact by using binding-deficient Pbx constructs to rescue a Pbx mutant fish line. A second question is why Meis, Pbx and Hox proteins are all required for transcription of Hox targets. We find that Meis, Pbx and Hox interact with transcriptional co-regulators (activators and repressors) and we will test if these interactions are essential in vivo. First we will use mutant forms of each protein that cannot interact with co-regulators and determine if these mutant forms are active. Second, we will interfere with co-regulator function in vivo and test if this affects expression of Hox target genes. Our results will provide information about hindbrain development and, since Hox proteins are involved in a number of aspects of neural development (e.g., dorsoventral patterning of the neural tube), will also be broadly applicable to neural development. hox, meis and pbx genes are also proto-oncogenes involved in leukemia and, by extrapolation, our results may also provide insights into the oncogenic nature of Hox proteins and their co-factors.